libreplanet-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Should Canoeboot become GNU Canoeboot?


From: Denis 'GNUtoo' Carikli
Subject: Re: Should Canoeboot become GNU Canoeboot?
Date: Sat, 18 May 2024 23:44:17 +0200

On Wed, 15 May 2024 18:00:12 +0200
Adrien 'neox' Bourmault <neox@gnu.org> wrote:
> FYI, in the first place, Denis and I tried to convince Leah to use
> the name osboot for her project and leave Libreboot to be the 100%
> libre project. She refused.
There is a lot of history and it's complicated to summarize it, so
I'll try to clarify it a bit.

For the record I did that by sending patches that started combining
both projects by making Libreboot a dependency of osboot, and still
have two separate projects. 

Beside reducing maintenance it also enabled more customization. The
patches worked but were more like an RFC as they didn't convert the full
projects, it was only the beginning.

More patches would have been made and sent later on if the patches sent
would have been accepted.

But they were refused later on with only a cryptic reason: being
against the new Libreboot policy, which If I remember well required
nonfree software to be shipped when the free version was technically
inferior or non-existent.

So I deduced that this was not a technical issue and not related to the
maintenance cost (the maintenance cost might have been a bit higher
with my approach than with directly shipping nonfree software in
Libreboot releases, though with my approach could potentially have bring
more contributions to Libreboot as it would have allowed more people to
write code to customize Libreboot and then clean up their code and
upstream it).

The precise real reason is still unknown to me, and I would still like
to know it but I can only make more or less vague suppositions
assumptions.

At the time after discussing during many hours with Leah, I understood
that move as being philosophically against fully free software because
it supposedly harmed free software, as it supposedly drove down
adoption, but I'm still not fully sure of that and I would still like
to understand more precisely why.

It could also not be to support more hardware with additional nonfree
software as my patches allowed to do that outside of Libreboot and that
explanation is incompatible with the reason of the refusal
(incompatible with Libreboot policy).

> We then took a bad decision : to use ourselves the name Libreboot (in
> the idea we could perhaps force her to stop using it).
More precisely, the issue was that a lot of people knew about Libreboot
being fully free and still as of today, many people I discuss with in
free software or political events don't know Libreboot stopped shipping
fully free binaries.

Because of that there are basically two approaches here:
- Reuse the same name and hope people would be confused and to clear
  the confusion they would look at both projects and understand what
  happened.

- Inform people and continue the original spirit of the project under
  a different name.

And we did both, and both have pros and cons. 

One compelling argument for the former is that I believe more in
(digital) commons than ownership from people, and so fighting for
preserving the association of a project name with the project values
looked like a good think since in case of issues most project usually
abandon their names without even a fight (Nexuiz, Gaim, etc), and what
matters most for me is not necessarily what project owners decide but to
respect the project values.

And I still had in mind that for a very long time people outside of the
free software community still referred to Open Office, and didn't know
about Libreoffice.

But at the end of the day someone convinced us to pick another name
because companies could sell 'Libreboot' laptops with a Libreboot
that contains nonfree software and still call it Libreboot.

So we picked a new name, and ironically Leah also pushed us to change
name to something like 'GNU Boot', so after evaluating many names we
finally picked this one.

Having users have the right information is important I think because I
believe that broader collaboration is only possible if people are
correctly informed. 

Because otherwise the less free projects shadow the more free ones as
people think they are 100% free whereas in reality they could
collaborate more: not having projects harm each other usually helps
collaboration a lot as people are more willing to collaborate when this
is the case. 

Even projects like Libreboot that now ship nonfree software can be
shadowed by projects with even more nonfree software: many computers
supported by Coreboot now use binaries from Intel or AMD that do almost
all of the work, and some computer vendors promote their laptops as
running Coreboot which is true, but in the head of many people Coreboot
is synonym with fully free software.

And even in the case of GNU Boot, on all the currently supported
non-laptops computers, I'd like to remind that in certain
configurations, plugging in a GPU results in the nonfree video BIOS of
the GPU being run by GNU Boot. Here the RYF certification helps
avoiding that kind of traps.

Also sometimes hardware also have builtin nonfree firmware like for
laptops keyboards, storage devices (HDD, SSD, microSD, etc), but here
we'd probably need more people in free software to tackle that more
broadly.

Anyway how or if different projects can collaborate also depends a lot
on the details and the context. For instance being able to install
extremely easily any nonfree software on top of any GNU/Linux
distributions makes it hard to resist to nonfree software when we are
pressured to use it, so that can threatens software freedom as a whole,
the development of alternatives to nonfree software, etc.

In the case of Libreboot, it also exists because of that kind of
collaboration: I worked on freeing the X60 GPU in Coreboot, with the
help of other Coreboot core developers. And to do that I adapted the
Coreboot GPU driver of the Chromebook Pixel that was supported in
Coreboot. This Chromebook was and is probably still a lost cause for
fully free software as to boot it required some code to be signed by
Intel. Later on people in Coreboot improved this driver and rewrote it
multiple time.

Libreboot could then exist because this free GPU driver enabled to make
I945 Thinkpads usable without the nonfree video BIOS, and also because
I gave Leah lot of explanations and the build scripts I used.

And now in some way GNU Boot even exists thanks to Libreboot. And if I
understood right, Canoeboot also exists thanks to GNU Boot in the same
strange way as it seems to have been created in reaction to GNU Boot
somehow.

> We need help to improve our documentation, indeed. But our project is
> definitely not dead. 
About that we're looking for help. We need people to help us review and
if necessary fix the installation instructions of GM45 Thinkpads like
the ThinkPad X200/X200s/X200T/T400/T500/W500/etc. We'd also like to
simplify the instructions as much as possible.

More specifically we're looking with people with flash chip programmers
that are easy to find, don't damage laptops if used correctly, and do
work with fully free software to help test and fix the instructions,
and adapt them to their programmer.

And here this is also something that could be useful for Canoeboot
depending on how Canoeboot wants its documentation to be (simple or
have many details).

Denis.

Attachment: pgpmvFQZGUZPm.pgp
Description: OpenPGP digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]